Abstract / Summary
Background and objectives Non-invasive ventilation (NIV) is increasingly used as a first-line therapy in acute hypoxemic respiratory failure (AHRF). Despite its benefits, NIV failure remains frequent. The HACOR score was initially developed to predict NIV failure in patients with AHRF. Later, the updated HACOR score (HACOR-U), which integrates baseline systemic severity with the HACOR score, was introduced to improve the prediction of NIV failure. This study aimed to validate the predictive accuracy of HACOR-U for NIV failure among patients with AHRF. Methods Fifty patients with AHRF initiated on NIV were enrolled in this study. HACOR and HACOR-U scores were calculated at baseline and again 1–2 h after NIV initiation. The predictive performance of both scores for NIV failure was assessed using area under the receiver operating characteristic curve (AUROC) analysis. 28-d mortality was also evaluated. Results The study population (n=50) comprised 58% men, with a mean age of 50 ± 15 y. Community-acquired pneumonia (34%) was the leading cause of AHRF. The NIV failure rate was 40%. HACOR-U demonstrated fair predictive accuracy for NIV failure, with the AUROC observed at 1–2 h (0.764), followed by baseline HACOR-U (0.745) and the 1–2-h HACOR score (0.649). At baseline, HACOR-U showed significantly better discrimination for NIV failure than HACOR (AUROC: 0.745 vs . 0.601, P =0.047). The overall 28-d mortality was 48% and was significantly higher among patients with NIV failure compared with NIV success (90% vs . 20%, P <0.001). Interpretation and conclusions In this validation study, HACOR-U demonstrated fair discriminatory ability for predicting NIV failure and may be useful for baseline risk stratification in patients with AHRF. However, larger studies are needed to confirm their clinical utility.